Pearl is the company that in 2021 received the first FDA clearance for AI software that reads dental radiographs — and has been systematically expanding coverage since: today it is the only one with US clearances for both 2D and 3D analysis. If you follow this tool category, Pearl is the reference point others are measured against.
Transparency note: this review is based on the manufacturer's official documentation and independent sources (listed at the bottom). I have no partnership with Pearl and earn no commission.
What Second Opinion is
Second Opinion is Pearl's flagship product: an AI that overlays color-coded annotations on the scan in real time, chairside. When you open a radiograph, flagged zones appear within seconds — while the patient is still in the chair.
It detects up to 18 finding types per image, including:
- suspected caries (at multiple stages),
- calculus,
- periapical radiolucencies,
- defective margins on existing restorations,
- bone loss (with automated level measurement),
- existing work: crowns, bridges, implants, fillings, root canals.
That last part — automatic recognition of existing work — also helps with records: AI clinical documentation is a topic of its own.
Regulatory status: why Pearl is special
The sequence of FDA clearances neatly shows where the whole category is going:
- 2021 — the first FDA clearance for AI reading of dental radiographs at all (bite-wing/periapical),
- May 2025 — automated bone level measurement + a pediatric indication,
- May 2025 — Second Opinion 3D for CBCT (first with both 2D and 3D clearance),
- December 2025 — detection on panoramic radiographs.
Beyond the US, the company reports regulatory clearance in 120 countries and use in over 23,000 practices, with integrations across more than 30 imaging and practice-management systems. Wherever you practice, confirm availability for your market and the data-processing terms (more below).
Where it helps day to day
A second pair of eyes in real time. Unlike tools that return a report afterwards, Second Opinion is designed to work while you examine — a principle I unpack in the article on AI and dental X-rays.
Patient communication. An annotated scan on the screen is the most convincing explanation tool you have: the patient sees what you see. That directly helps case acceptance.
Team consistency. The same criteria for every team member, every scan, every day.
What it costs
No public price list — subscription pricing set through a sales conversation (practice size, number of locations). Before the demo call, I recommend running the savings calculator and defining what the tool must return to be worth it.
What to watch out for
- Scans are processed in the cloud — check where they're stored, whether they're used for training, and the data-processing agreement; the question list is in patient privacy and AI.
- Detection is not diagnosis — the system flags suspicions; the clinical decision is yours. False positives exist here too.
- Integration with your software — 30+ integrations sounds good, but verify your exact sensor/program combination before signing.
Pearl or Diagnocat — or something else?
In short: Pearl is strongest in real-time 2D detection chairside and breadth of regulatory clearances; Diagnocat stands out in CBCT segmentation and implant planning; Overjet targets large groups and insurers. Start from what you do most — then look at demos. I keep an overview of the whole category on the AI tools page.
In short
- The first FDA-cleared AI for dental radiographs; today the only one cleared for 2D + 3D + panoramic.
- Up to 18 finding types, in real time, with bone level measurement.
- 23,000+ practices, clearance in 120 countries, 30+ integrations.
- Pricing on request — prepare your ROI math in advance.
- You verify the findings: AI is decision support, not the decision maker.
For the bigger picture, see the complete guide to AI in dentistry.
